1 /* 2 * Copyright (C) 2014 The Android Open Source Project 3 * Copyright (c) 1999, 2013, Oracle and/or its affiliates. All rights reserved. 4 * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. 5 * 6 * This code is free software; you can redistribute it and/or modify it 7 * under the terms of the GNU General Public License version 2 only, as 8 * published by the Free Software Foundation. Oracle designates this 9 * particular file as subject to the "Classpath" exception as provided 10 * by Oracle in the LICENSE file that accompanied this code. 11 * 12 * This code is distributed in the hope that it will be useful, but WITHOUT 13 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or 14 * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License 15 * version 2 for more details (a copy is included in the LICENSE file that 16 * accompanied this code). 17 * 18 * You should have received a copy of the GNU General Public License version 19 * 2 along with this work; if not, write to the Free Software Foundation, 20 * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. 21 * 22 * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA 23 * or visit www.oracle.com if you need additional information or have any 24 * questions. 25 */ 26 27 package java.util.regex; 28 29 import dalvik.annotation.optimization.ReachabilitySensitive; 30 import dalvik.system.VMRuntime; 31 32 import libcore.util.NativeAllocationRegistry; 33 34 import java.util.Iterator; 35 import java.util.ArrayList; 36 import java.util.NoSuchElementException; 37 import java.util.Spliterator; 38 import java.util.Spliterators; 39 import java.util.function.Predicate; 40 import java.util.stream.Stream; 41 import java.util.stream.StreamSupport; 42 43 import libcore.util.EmptyArray; 44 45 // Android-changed: Document that named capturing is only available from API 26. 46 // Android-changed: Android always uses unicode character classes. 47 // UNICODE_CHARACTER_CLASS has no effect on Android. 48 /** 49 * A compiled representation of a regular expression. 50 * 51 * <p> A regular expression, specified as a string, must first be compiled into 52 * an instance of this class. The resulting pattern can then be used to create 53 * a {@link Matcher} object that can match arbitrary {@linkplain 54 * java.lang.CharSequence character sequences} against the regular 55 * expression. All of the state involved in performing a match resides in the 56 * matcher, so many matchers can share the same pattern. 57 * 58 * <p> A typical invocation sequence is thus 59 * 60 * <blockquote><pre> 61 * Pattern p = Pattern.{@link #compile compile}("a*b"); 62 * Matcher m = p.{@link #matcher matcher}("aaaaab"); 63 * boolean b = m.{@link Matcher#matches matches}();</pre></blockquote> 64 * 65 * <p> A {@link #matches matches} method is defined by this class as a 66 * convenience for when a regular expression is used just once. This method 67 * compiles an expression and matches an input sequence against it in a single 68 * invocation. The statement 69 * 70 * <blockquote><pre> 71 * boolean b = Pattern.matches("a*b", "aaaaab");</pre></blockquote> 72 * 73 * is equivalent to the three statements above, though for repeated matches it 74 * is less efficient since it does not allow the compiled pattern to be reused. 75 * 76 * <p> Instances of this class are immutable and are safe for use by multiple 77 * concurrent threads. Instances of the {@link Matcher} class are not safe for 78 * such use. 79 * 80 * 81 * <h3><a name="sum">Summary of regular-expression constructs</a></h3> 82 * 83 * <table border="0" cellpadding="1" cellspacing="0" 84 * summary="Regular expression constructs, and what they match"> 85 * 86 * <tr align="left"> 87 * <th align="left" id="construct">Construct</th> 88 * <th align="left" id="matches">Matches</th> 89 * </tr> 90 * 91 * <tr><th> </th></tr> 92 * <tr align="left"><th colspan="2" id="characters">Characters</th></tr> 93 * 94 * <tr><td valign="top" headers="construct characters"><i>x</i></td> 95 * <td headers="matches">The character <i>x</i></td></tr> 96 * <tr><td valign="top" headers="construct characters"><tt>\\</tt></td> 97 * <td headers="matches">The backslash character</td></tr> 98 * <tr><td valign="top" headers="construct characters"><tt>\0</tt><i>n</i></td> 99 * <td headers="matches">The character with octal value <tt>0</tt><i>n</i> 100 * (0 <tt><=</tt> <i>n</i> <tt><=</tt> 7)</td></tr> 101 * <tr><td valign="top" headers="construct characters"><tt>\0</tt><i>nn</i></td> 102 * <td headers="matches">The character with octal value <tt>0</tt><i>nn</i> 103 * (0 <tt><=</tt> <i>n</i> <tt><=</tt> 7)</td></tr> 104 * <tr><td valign="top" headers="construct characters"><tt>\0</tt><i>mnn</i></td> 105 * <td headers="matches">The character with octal value <tt>0</tt><i>mnn</i> 106 * (0 <tt><=</tt> <i>m</i> <tt><=</tt> 3, 107 * 0 <tt><=</tt> <i>n</i> <tt><=</tt> 7)</td></tr> 108 * <tr><td valign="top" headers="construct characters"><tt>\x</tt><i>hh</i></td> 109 * <td headers="matches">The character with hexadecimal value <tt>0x</tt><i>hh</i></td></tr> 110 * <tr><td valign="top" headers="construct characters"><tt>\u</tt><i>hhhh</i></td> 111 * <td headers="matches">The character with hexadecimal value <tt>0x</tt><i>hhhh</i></td></tr> 112 * <tr><td valign="top" headers="construct characters"><tt>\x</tt><i>{h...h}</i></td> 113 * <td headers="matches">The character with hexadecimal value <tt>0x</tt><i>h...h</i> 114 * ({@link java.lang.Character#MIN_CODE_POINT Character.MIN_CODE_POINT} 115 * <= <tt>0x</tt><i>h...h</i> <= 116 * {@link java.lang.Character#MAX_CODE_POINT Character.MAX_CODE_POINT})</td></tr> 117 * <tr><td valign="top" headers="matches"><tt>\t</tt></td> 118 * <td headers="matches">The tab character (<tt>'\u0009'</tt>)</td></tr> 119 * <tr><td valign="top" headers="construct characters"><tt>\n</tt></td> 120 * <td headers="matches">The newline (line feed) character (<tt>'\u000A'</tt>)</td></tr> 121 * <tr><td valign="top" headers="construct characters"><tt>\r</tt></td> 122 * <td headers="matches">The carriage-return character (<tt>'\u000D'</tt>)</td></tr> 123 * <tr><td valign="top" headers="construct characters"><tt>\f</tt></td> 124 * <td headers="matches">The form-feed character (<tt>'\u000C'</tt>)</td></tr> 125 * <tr><td valign="top" headers="construct characters"><tt>\a</tt></td> 126 * <td headers="matches">The alert (bell) character (<tt>'\u0007'</tt>)</td></tr> 127 * <tr><td valign="top" headers="construct characters"><tt>\e</tt></td> 128 * <td headers="matches">The escape character (<tt>'\u001B'</tt>)</td></tr> 129 * <tr><td valign="top" headers="construct characters"><tt>\c</tt><i>x</i></td> 130 * <td headers="matches">The control character corresponding to <i>x</i></td></tr> 131 * 132 * <tr><th> </th></tr> 133 * <tr align="left"><th colspan="2" id="classes">Character classes</th></tr> 134 * 135 * <tr><td valign="top" headers="construct classes">{@code [abc]}</td> 136 * <td headers="matches">{@code a}, {@code b}, or {@code c} (simple class)</td></tr> 137 * <tr><td valign="top" headers="construct classes">{@code [^abc]}</td> 138 * <td headers="matches">Any character except {@code a}, {@code b}, or {@code c} (negation)</td></tr> 139 * <tr><td valign="top" headers="construct classes">{@code [a-zA-Z]}</td> 140 * <td headers="matches">{@code a} through {@code z} 141 * or {@code A} through {@code Z}, inclusive (range)</td></tr> 142 * <tr><td valign="top" headers="construct classes">{@code [a-d[m-p]]}</td> 143 * <td headers="matches">{@code a} through {@code d}, 144 * or {@code m} through {@code p}: {@code [a-dm-p]} (union)</td></tr> 145 * <tr><td valign="top" headers="construct classes">{@code [a-z&&[def]]}</td> 146 * <td headers="matches">{@code d}, {@code e}, or {@code f} (intersection)</tr> 147 * <tr><td valign="top" headers="construct classes">{@code [a-z&&[^bc]]}</td> 148 * <td headers="matches">{@code a} through {@code z}, 149 * except for {@code b} and {@code c}: {@code [ad-z]} (subtraction)</td></tr> 150 * <tr><td valign="top" headers="construct classes">{@code [a-z&&[^m-p]]}</td> 151 * <td headers="matches">{@code a} through {@code z}, 152 * and not {@code m} through {@code p}: {@code [a-lq-z]}(subtraction)</td></tr> 153 * <tr><th> </th></tr> 154 * 155 * <tr align="left"><th colspan="2" id="predef">Predefined character classes</th></tr> 156 * 157 * <tr><td valign="top" headers="construct predef"><tt>.</tt></td> 158 * <td headers="matches">Any character (may or may not match <a href="#lt">line terminators</a>)</td></tr> 159 * <tr><td valign="top" headers="construct predef"><tt>\d</tt></td> 160 * <td headers="matches">A digit: <tt>[0-9]</tt></td></tr> 161 * <tr><td valign="top" headers="construct predef"><tt>\D</tt></td> 162 * <td headers="matches">A non-digit: <tt>[^0-9]</tt></td></tr> 163 * <tr><td valign="top" headers="construct predef"><tt>\h</tt></td> 164 * <td headers="matches">A horizontal whitespace character: 165 * <tt>[ \t\xA0\u1680\u180e\u2000-\u200a\u202f\u205f\u3000]</tt></td></tr> 166 * <tr><td valign="top" headers="construct predef"><tt>\H</tt></td> 167 * <td headers="matches">A non-horizontal whitespace character: <tt>[^\h]</tt></td></tr> 168 * <tr><td valign="top" headers="construct predef"><tt>\s</tt></td> 169 * <td headers="matches">A whitespace character: <tt>[ \t\n\x0B\f\r]</tt></td></tr> 170 * <tr><td valign="top" headers="construct predef"><tt>\S</tt></td> 171 * <td headers="matches">A non-whitespace character: <tt>[^\s]</tt></td></tr> 172 * <tr><td valign="top" headers="construct predef"><tt>\v</tt></td> 173 * <td headers="matches">A vertical whitespace character: <tt>[\n\x0B\f\r\x85\u2028\u2029]</tt> 174 * </td></tr> 175 * <tr><td valign="top" headers="construct predef"><tt>\V</tt></td> 176 * <td headers="matches">A non-vertical whitespace character: <tt>[^\v]</tt></td></tr> 177 * <tr><td valign="top" headers="construct predef"><tt>\w</tt></td> 178 * <td headers="matches">A word character: <tt>[a-zA-Z_0-9]</tt></td></tr> 179 * <tr><td valign="top" headers="construct predef"><tt>\W</tt></td> 180 * <td headers="matches">A non-word character: <tt>[^\w]</tt></td></tr> 181 * <tr><th> </th></tr> 182 * <tr align="left"><th colspan="2" id="posix"><b>POSIX character classes (US-ASCII only)</b></th></tr> 183 * 184 * <tr><td valign="top" headers="construct posix">{@code \p{Lower}}</td> 185 * <td headers="matches">A lower-case alphabetic character: {@code [a-z]}</td></tr> 186 * <tr><td valign="top" headers="construct posix">{@code \p{Upper}}</td> 187 * <td headers="matches">An upper-case alphabetic character:{@code [A-Z]}</td></tr> 188 * <tr><td valign="top" headers="construct posix">{@code \p{ASCII}}</td> 189 * <td headers="matches">All ASCII:{@code [\x00-\x7F]}</td></tr> 190 * <tr><td valign="top" headers="construct posix">{@code \p{Alpha}}</td> 191 * <td headers="matches">An alphabetic character:{@code [\p{Lower}\p{Upper}]}</td></tr> 192 * <tr><td valign="top" headers="construct posix">{@code \p{Digit}}</td> 193 * <td headers="matches">A decimal digit: {@code [0-9]}</td></tr> 194 * <tr><td valign="top" headers="construct posix">{@code \p{Alnum}}</td> 195 * <td headers="matches">An alphanumeric character:{@code [\p{Alpha}\p{Digit}]}</td></tr> 196 * <tr><td valign="top" headers="construct posix">{@code \p{Punct}}</td> 197 * <td headers="matches">Punctuation: One of {@code !"#$%&'()*+,-./:;<=>?@[\]^_`{|}~}</td></tr> 198 * <!-- {@code [\!"#\$%&'\(\)\*\+,\-\./:;\<=\>\?@\[\\\]\^_`\{\|\}~]} 199 * {@code [\X21-\X2F\X31-\X40\X5B-\X60\X7B-\X7E]} --> 200 * <tr><td valign="top" headers="construct posix">{@code \p{Graph}}</td> 201 * <td headers="matches">A visible character: {@code [\p{Alnum}\p{Punct}]}</td></tr> 202 * <tr><td valign="top" headers="construct posix">{@code \p{Print}}</td> 203 * <td headers="matches">A printable character: {@code [\p{Graph}\x20]}</td></tr> 204 * <tr><td valign="top" headers="construct posix">{@code \p{Blank}}</td> 205 * <td headers="matches">A space or a tab: {@code [ \t]}</td></tr> 206 * <tr><td valign="top" headers="construct posix">{@code \p{Cntrl}}</td> 207 * <td headers="matches">A control character: {@code [\x00-\x1F\x7F]}</td></tr> 208 * <tr><td valign="top" headers="construct posix">{@code \p{XDigit}}</td> 209 * <td headers="matches">A hexadecimal digit: {@code [0-9a-fA-F]}</td></tr> 210 * <tr><td valign="top" headers="construct posix">{@code \p{Space}}</td> 211 * <td headers="matches">A whitespace character: {@code [ \t\n\x0B\f\r]}</td></tr> 212 * 213 * <tr><th> </th></tr> 214 * <tr align="left"><th colspan="2">java.lang.Character classes (simple <a href="#jcc">java character type</a>)</th></tr> 215 * 216 * <tr><td valign="top"><tt>\p{javaLowerCase}</tt></td> 217 * <td>Equivalent to java.lang.Character.isLowerCase()</td></tr> 218 * <tr><td valign="top"><tt>\p{javaUpperCase}</tt></td> 219 * <td>Equivalent to java.lang.Character.isUpperCase()</td></tr> 220 * <tr><td valign="top"><tt>\p{javaWhitespace}</tt></td> 221 * <td>Equivalent to java.lang.Character.isWhitespace()</td></tr> 222 * <tr><td valign="top"><tt>\p{javaMirrored}</tt></td> 223 * <td>Equivalent to java.lang.Character.isMirrored()</td></tr> 224 * 225 * <tr><th> </th></tr> 226 * <tr align="left"><th colspan="2" id="unicode">Classes for Unicode scripts, blocks, categories and binary properties</th></tr> 227 * <tr><td valign="top" headers="construct unicode">{@code \p{IsLatin}}</td> 228 * <td headers="matches">A Latin script character (<a href="#usc">script</a>)</td></tr> 229 * <tr><td valign="top" headers="construct unicode">{@code \p{InGreek}}</td> 230 * <td headers="matches">A character in the Greek block (<a href="#ubc">block</a>)</td></tr> 231 * <tr><td valign="top" headers="construct unicode">{@code \p{Lu}}</td> 232 * <td headers="matches">An uppercase letter (<a href="#ucc">category</a>)</td></tr> 233 * <tr><td valign="top" headers="construct unicode">{@code \p{IsAlphabetic}}</td> 234 * <td headers="matches">An alphabetic character (<a href="#ubpc">binary property</a>)</td></tr> 235 * <tr><td valign="top" headers="construct unicode">{@code \p{Sc}}</td> 236 * <td headers="matches">A currency symbol</td></tr> 237 * <tr><td valign="top" headers="construct unicode">{@code \P{InGreek}}</td> 238 * <td headers="matches">Any character except one in the Greek block (negation)</td></tr> 239 * <tr><td valign="top" headers="construct unicode">{@code [\p{L}&&[^\p{Lu}]]}</td> 240 * <td headers="matches">Any letter except an uppercase letter (subtraction)</td></tr> 241 * 242 * <tr><th> </th></tr> 243 * <tr align="left"><th colspan="2" id="bounds">Boundary matchers</th></tr> 244 * 245 * <tr><td valign="top" headers="construct bounds"><tt>^</tt></td> 246 * <td headers="matches">The beginning of a line</td></tr> 247 * <tr><td valign="top" headers="construct bounds"><tt>$</tt></td> 248 * <td headers="matches">The end of a line</td></tr> 249 * <tr><td valign="top" headers="construct bounds"><tt>\b</tt></td> 250 * <td headers="matches">A word boundary</td></tr> 251 * <tr><td valign="top" headers="construct bounds"><tt>\B</tt></td> 252 * <td headers="matches">A non-word boundary</td></tr> 253 * <tr><td valign="top" headers="construct bounds"><tt>\A</tt></td> 254 * <td headers="matches">The beginning of the input</td></tr> 255 * <tr><td valign="top" headers="construct bounds"><tt>\G</tt></td> 256 * <td headers="matches">The end of the previous match</td></tr> 257 * <tr><td valign="top" headers="construct bounds"><tt>\Z</tt></td> 258 * <td headers="matches">The end of the input but for the final 259 * <a href="#lt">terminator</a>, if any</td></tr> 260 * <tr><td valign="top" headers="construct bounds"><tt>\z</tt></td> 261 * <td headers="matches">The end of the input</td></tr> 262 * 263 * <tr><th> </th></tr> 264 * <tr align="left"><th colspan="2" id="lineending">Linebreak matcher</th></tr> 265 * <tr><td valign="top" headers="construct lineending"><tt>\R</tt></td> 266 * <td headers="matches">Any Unicode linebreak sequence, is equivalent to 267 * <tt>\u000D\u000A|[\u000A\u000B\u000C\u000D\u0085\u2028\u2029] 268 * </tt></td></tr> 269 * 270 * <tr><th> </th></tr> 271 * <tr align="left"><th colspan="2" id="greedy">Greedy quantifiers</th></tr> 272 * 273 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>?</tt></td> 274 * <td headers="matches"><i>X</i>, once or not at all</td></tr> 275 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>*</tt></td> 276 * <td headers="matches"><i>X</i>, zero or more times</td></tr> 277 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>+</tt></td> 278 * <td headers="matches"><i>X</i>, one or more times</td></tr> 279 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>{</tt><i>n</i><tt>}</tt></td> 280 * <td headers="matches"><i>X</i>, exactly <i>n</i> times</td></tr> 281 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>{</tt><i>n</i><tt>,}</tt></td> 282 * <td headers="matches"><i>X</i>, at least <i>n</i> times</td></tr> 283 * <tr><td valign="top" headers="construct greedy"><i>X</i><tt>{</tt><i>n</i><tt>,</tt><i>m</i><tt>}</tt></td> 284 * <td headers="matches"><i>X</i>, at least <i>n</i> but not more than <i>m</i> times</td></tr> 285 * 286 * <tr><th> </th></tr> 287 * <tr align="left"><th colspan="2" id="reluc">Reluctant quantifiers</th></tr> 288 * 289 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>??</tt></td> 290 * <td headers="matches"><i>X</i>, once or not at all</td></tr> 291 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>*?</tt></td> 292 * <td headers="matches"><i>X</i>, zero or more times</td></tr> 293 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>+?</tt></td> 294 * <td headers="matches"><i>X</i>, one or more times</td></tr> 295 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>{</tt><i>n</i><tt>}?</tt></td> 296 * <td headers="matches"><i>X</i>, exactly <i>n</i> times</td></tr> 297 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>{</tt><i>n</i><tt>,}?</tt></td> 298 * <td headers="matches"><i>X</i>, at least <i>n</i> times</td></tr> 299 * <tr><td valign="top" headers="construct reluc"><i>X</i><tt>{</tt><i>n</i><tt>,</tt><i>m</i><tt>}?</tt></td> 300 * <td headers="matches"><i>X</i>, at least <i>n</i> but not more than <i>m</i> times</td></tr> 301 * 302 * <tr><th> </th></tr> 303 * <tr align="left"><th colspan="2" id="poss">Possessive quantifiers</th></tr> 304 * 305 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>?+</tt></td> 306 * <td headers="matches"><i>X</i>, once or not at all</td></tr> 307 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>*+</tt></td> 308 * <td headers="matches"><i>X</i>, zero or more times</td></tr> 309 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>++</tt></td> 310 * <td headers="matches"><i>X</i>, one or more times</td></tr> 311 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>{</tt><i>n</i><tt>}+</tt></td> 312 * <td headers="matches"><i>X</i>, exactly <i>n</i> times</td></tr> 313 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>{</tt><i>n</i><tt>,}+</tt></td> 314 * <td headers="matches"><i>X</i>, at least <i>n</i> times</td></tr> 315 * <tr><td valign="top" headers="construct poss"><i>X</i><tt>{</tt><i>n</i><tt>,</tt><i>m</i><tt>}+</tt></td> 316 * <td headers="matches"><i>X</i>, at least <i>n</i> but not more than <i>m</i> times</td></tr> 317 * 318 * <tr><th> </th></tr> 319 * <tr align="left"><th colspan="2" id="logical">Logical operators</th></tr> 320 * 321 * <tr><td valign="top" headers="construct logical"><i>XY</i></td> 322 * <td headers="matches"><i>X</i> followed by <i>Y</i></td></tr> 323 * <tr><td valign="top" headers="construct logical"><i>X</i><tt>|</tt><i>Y</i></td> 324 * <td headers="matches">Either <i>X</i> or <i>Y</i></td></tr> 325 * <tr><td valign="top" headers="construct logical"><tt>(</tt><i>X</i><tt>)</tt></td> 326 * <td headers="matches">X, as a <a href="#cg">capturing group</a></td></tr> 327 * 328 * <tr><th> </th></tr> 329 * <tr align="left"><th colspan="2" id="backref">Back references</th></tr> 330 * 331 * <tr><td valign="bottom" headers="construct backref"><tt>\</tt><i>n</i></td> 332 * <td valign="bottom" headers="matches">Whatever the <i>n</i><sup>th</sup> 333 * <a href="#cg">capturing group</a> matched</td></tr> 334 * 335 * <tr><td valign="bottom" headers="construct backref"><tt>\</tt><i>k</i><<i>name</i>></td> 336 * <td valign="bottom" headers="matches">Whatever the 337 * <a href="#groupname">named-capturing group</a> "name" matched. Only available for API 26 or above</td></tr> 338 * 339 * <tr><th> </th></tr> 340 * <tr align="left"><th colspan="2" id="quot">Quotation</th></tr> 341 * 342 * <tr><td valign="top" headers="construct quot"><tt>\</tt></td> 343 * <td headers="matches">Nothing, but quotes the following character</td></tr> 344 * <tr><td valign="top" headers="construct quot"><tt>\Q</tt></td> 345 * <td headers="matches">Nothing, but quotes all characters until <tt>\E</tt></td></tr> 346 * <tr><td valign="top" headers="construct quot"><tt>\E</tt></td> 347 * <td headers="matches">Nothing, but ends quoting started by <tt>\Q</tt></td></tr> 348 * <!-- Metachars: !$()*+.<>?[\]^{|} --> 349 * 350 * <tr><th> </th></tr> 351 * <tr align="left"><th colspan="2" id="special">Special constructs (named-capturing and non-capturing)</th></tr> 352 * 353 * <tr><td valign="top" headers="construct special"><tt>(?<<a href="#groupname">name</a>></tt><i>X</i><tt>)</tt></td> 354 * <td headers="matches"><i>X</i>, as a named-capturing group. Only available for API 26 or above.</td></tr> 355 * <tr><td valign="top" headers="construct special"><tt>(?:</tt><i>X</i><tt>)</tt></td> 356 * <td headers="matches"><i>X</i>, as a non-capturing group</td></tr> 357 * <tr><td valign="top" headers="construct special"><tt>(?idmsuxU-idmsuxU) </tt></td> 358 * <td headers="matches">Nothing, but turns match flags <a href="#CASE_INSENSITIVE">i</a> 359 * <a href="#UNIX_LINES">d</a> <a href="#MULTILINE">m</a> <a href="#DOTALL">s</a> 360 * <a href="#UNICODE_CASE">u</a> <a href="#COMMENTS">x</a> <a href="#UNICODE_CHARACTER_CLASS">U</a> 361 * on - off</td></tr> 362 * <tr><td valign="top" headers="construct special"><tt>(?idmsux-idmsux:</tt><i>X</i><tt>)</tt> </td> 363 * <td headers="matches"><i>X</i>, as a <a href="#cg">non-capturing group</a> with the 364 * given flags <a href="#CASE_INSENSITIVE">i</a> <a href="#UNIX_LINES">d</a> 365 * <a href="#MULTILINE">m</a> <a href="#DOTALL">s</a> <a href="#UNICODE_CASE">u</a > 366 * <a href="#COMMENTS">x</a> on - off</td></tr> 367 * <tr><td valign="top" headers="construct special"><tt>(?=</tt><i>X</i><tt>)</tt></td> 368 * <td headers="matches"><i>X</i>, via zero-width positive lookahead</td></tr> 369 * <tr><td valign="top" headers="construct special"><tt>(?!</tt><i>X</i><tt>)</tt></td> 370 * <td headers="matches"><i>X</i>, via zero-width negative lookahead</td></tr> 371 * <tr><td valign="top" headers="construct special"><tt>(?<=</tt><i>X</i><tt>)</tt></td> 372 * <td headers="matches"><i>X</i>, via zero-width positive lookbehind</td></tr> 373 * <tr><td valign="top" headers="construct special"><tt>(?<!</tt><i>X</i><tt>)</tt></td> 374 * <td headers="matches"><i>X</i>, via zero-width negative lookbehind</td></tr> 375 * <tr><td valign="top" headers="construct special"><tt>(?></tt><i>X</i><tt>)</tt></td> 376 * <td headers="matches"><i>X</i>, as an independent, non-capturing group</td></tr> 377 * 378 * </table> 379 * 380 * <hr> 381 * 382 * 383 * <h3><a name="bs">Backslashes, escapes, and quoting</a></h3> 384 * 385 * <p> The backslash character (<tt>'\'</tt>) serves to introduce escaped 386 * constructs, as defined in the table above, as well as to quote characters 387 * that otherwise would be interpreted as unescaped constructs. Thus the 388 * expression <tt>\\</tt> matches a single backslash and <tt>\{</tt> matches a 389 * left brace. 390 * 391 * <p> It is an error to use a backslash prior to any alphabetic character that 392 * does not denote an escaped construct; these are reserved for future 393 * extensions to the regular-expression language. A backslash may be used 394 * prior to a non-alphabetic character regardless of whether that character is 395 * part of an unescaped construct. 396 * 397 * <p> Backslashes within string literals in Java source code are interpreted 398 * as required by 399 * <cite>The Java™ Language Specification</cite> 400 * as either Unicode escapes (section 3.3) or other character escapes (section 3.10.6) 401 * It is therefore necessary to double backslashes in string 402 * literals that represent regular expressions to protect them from 403 * interpretation by the Java bytecode compiler. The string literal 404 * <tt>"\b"</tt>, for example, matches a single backspace character when 405 * interpreted as a regular expression, while <tt>"\\b"</tt> matches a 406 * word boundary. The string literal <tt>"\(hello\)"</tt> is illegal 407 * and leads to a compile-time error; in order to match the string 408 * <tt>(hello)</tt> the string literal <tt>"\\(hello\\)"</tt> 409 * must be used. 410 * 411 * <h3><a name="cc">Character Classes</a></h3> 412 * 413 * <p> Character classes may appear within other character classes, and 414 * may be composed by the union operator (implicit) and the intersection 415 * operator (<tt>&&</tt>). 416 * The union operator denotes a class that contains every character that is 417 * in at least one of its operand classes. The intersection operator 418 * denotes a class that contains every character that is in both of its 419 * operand classes. 420 * 421 * <p> The precedence of character-class operators is as follows, from 422 * highest to lowest: 423 * 424 * <blockquote><table border="0" cellpadding="1" cellspacing="0" 425 * summary="Precedence of character class operators."> 426 * <tr><th>1 </th> 427 * <td>Literal escape </td> 428 * <td><tt>\x</tt></td></tr> 429 * <tr><th>2 </th> 430 * <td>Grouping</td> 431 * <td><tt>[...]</tt></td></tr> 432 * <tr><th>3 </th> 433 * <td>Range</td> 434 * <td><tt>a-z</tt></td></tr> 435 * <tr><th>4 </th> 436 * <td>Union</td> 437 * <td><tt>[a-e][i-u]</tt></td></tr> 438 * <tr><th>5 </th> 439 * <td>Intersection</td> 440 * <td>{@code [a-z&&[aeiou]]}</td></tr> 441 * </table></blockquote> 442 * 443 * <p> Note that a different set of metacharacters are in effect inside 444 * a character class than outside a character class. For instance, the 445 * regular expression <tt>.</tt> loses its special meaning inside a 446 * character class, while the expression <tt>-</tt> becomes a range 447 * forming metacharacter. 448 * 449 * <h3><a name="lt">Line terminators</a></h3> 450 * 451 * <p> A <i>line terminator</i> is a one- or two-character sequence that marks 452 * the end of a line of the input character sequence. The following are 453 * recognized as line terminators: 454 * 455 * <ul> 456 * 457 * <li> A newline (line feed) character (<tt>'\n'</tt>), 458 * 459 * <li> A carriage-return character followed immediately by a newline 460 * character (<tt>"\r\n"</tt>), 461 * 462 * <li> A standalone carriage-return character (<tt>'\r'</tt>), 463 * 464 * <li> A next-line character (<tt>'\u0085'</tt>), 465 * 466 * <li> A line-separator character (<tt>'\u2028'</tt>), or 467 * 468 * <li> A paragraph-separator character (<tt>'\u2029</tt>). 469 * 470 * </ul> 471 * <p>If {@link #UNIX_LINES} mode is activated, then the only line terminators 472 * recognized are newline characters. 473 * 474 * <p> The regular expression <tt>.</tt> matches any character except a line 475 * terminator unless the {@link #DOTALL} flag is specified. 476 * 477 * <p> By default, the regular expressions <tt>^</tt> and <tt>$</tt> ignore 478 * line terminators and only match at the beginning and the end, respectively, 479 * of the entire input sequence. If {@link #MULTILINE} mode is activated then 480 * <tt>^</tt> matches at the beginning of input and after any line terminator 481 * except at the end of input. When in {@link #MULTILINE} mode <tt>$</tt> 482 * matches just before a line terminator or the end of the input sequence. 483 * 484 * <h3><a name="cg">Groups and capturing</a></h3> 485 * 486 * <h4><a name="gnumber">Group number</a></h4> 487 * <p> Capturing groups are numbered by counting their opening parentheses from 488 * left to right. In the expression <tt>((A)(B(C)))</tt>, for example, there 489 * are four such groups: </p> 490 * 491 * <blockquote><table cellpadding=1 cellspacing=0 summary="Capturing group numberings"> 492 * <tr><th>1 </th> 493 * <td><tt>((A)(B(C)))</tt></td></tr> 494 * <tr><th>2 </th> 495 * <td><tt>(A)</tt></td></tr> 496 * <tr><th>3 </th> 497 * <td><tt>(B(C))</tt></td></tr> 498 * <tr><th>4 </th> 499 * <td><tt>(C)</tt></td></tr> 500 * </table></blockquote> 501 * 502 * <p> Group zero always stands for the entire expression. 503 * 504 * <p> Capturing groups are so named because, during a match, each subsequence 505 * of the input sequence that matches such a group is saved. The captured 506 * subsequence may be used later in the expression, via a back reference, and 507 * may also be retrieved from the matcher once the match operation is complete. 508 * 509 * <h4><a name="groupname">Group name</a></h4> 510 * <p>The constructs and APIs are available since API level 26. A capturing group 511 * can also be assigned a "name", a <tt>named-capturing group</tt>, 512 * and then be back-referenced later by the "name". Group names are composed of 513 * the following characters. The first character must be a <tt>letter</tt>. 514 * 515 * <ul> 516 * <li> The uppercase letters <tt>'A'</tt> through <tt>'Z'</tt> 517 * (<tt>'\u0041'</tt> through <tt>'\u005a'</tt>), 518 * <li> The lowercase letters <tt>'a'</tt> through <tt>'z'</tt> 519 * (<tt>'\u0061'</tt> through <tt>'\u007a'</tt>), 520 * <li> The digits <tt>'0'</tt> through <tt>'9'</tt> 521 * (<tt>'\u0030'</tt> through <tt>'\u0039'</tt>), 522 * </ul> 523 * 524 * <p> A <tt>named-capturing group</tt> is still numbered as described in 525 * <a href="#gnumber">Group number</a>. 526 * 527 * <p> The captured input associated with a group is always the subsequence 528 * that the group most recently matched. If a group is evaluated a second time 529 * because of quantification then its previously-captured value, if any, will 530 * be retained if the second evaluation fails. Matching the string 531 * <tt>"aba"</tt> against the expression <tt>(a(b)?)+</tt>, for example, leaves 532 * group two set to <tt>"b"</tt>. All captured input is discarded at the 533 * beginning of each match. 534 * 535 * <p> Groups beginning with <tt>(?</tt> are either pure, <i>non-capturing</i> groups 536 * that do not capture text and do not count towards the group total, or 537 * <i>named-capturing</i> group. 538 * 539 * <h3> Unicode support </h3> 540 * 541 * <p> This class is in conformance with Level 1 of <a 542 * href="http://www.unicode.org/reports/tr18/"><i>Unicode Technical 543 * Standard #18: Unicode Regular Expression</i></a>, plus RL2.1 544 * Canonical Equivalents. 545 * <p> 546 * <b>Unicode escape sequences</b> such as <tt>\u2014</tt> in Java source code 547 * are processed as described in section 3.3 of 548 * <cite>The Java™ Language Specification</cite>. 549 * Such escape sequences are also implemented directly by the regular-expression 550 * parser so that Unicode escapes can be used in expressions that are read from 551 * files or from the keyboard. Thus the strings <tt>"\u2014"</tt> and 552 * <tt>"\\u2014"</tt>, while not equal, compile into the same pattern, which 553 * matches the character with hexadecimal value <tt>0x2014</tt>. 554 * <p> 555 * A Unicode character can also be represented in a regular-expression by 556 * using its <b>Hex notation</b>(hexadecimal code point value) directly as described in construct 557 * <tt>\x{...}</tt>, for example a supplementary character U+2011F 558 * can be specified as <tt>\x{2011F}</tt>, instead of two consecutive 559 * Unicode escape sequences of the surrogate pair 560 * <tt>\uD840</tt><tt>\uDD1F</tt>. 561 * <p> 562 * Unicode scripts, blocks, categories and binary properties are written with 563 * the <tt>\p</tt> and <tt>\P</tt> constructs as in Perl. 564 * <tt>\p{</tt><i>prop</i><tt>}</tt> matches if 565 * the input has the property <i>prop</i>, while <tt>\P{</tt><i>prop</i><tt>}</tt> 566 * does not match if the input has that property. 567 * <p> 568 * Scripts, blocks, categories and binary properties can be used both inside 569 * and outside of a character class. 570 * 571 * <p> 572 * <b><a name="usc">Scripts</a></b> are specified either with the prefix {@code Is}, as in 573 * {@code IsHiragana}, or by using the {@code script} keyword (or its short 574 * form {@code sc})as in {@code script=Hiragana} or {@code sc=Hiragana}. 575 * <p> 576 * The script names supported by <code>Pattern</code> are the valid script names 577 * accepted and defined by 578 * {@link java.lang.Character.UnicodeScript#forName(String) UnicodeScript.forName}. 579 * 580 * <p> 581 * <b><a name="ubc">Blocks</a></b> are specified with the prefix {@code In}, as in 582 * {@code InMongolian}, or by using the keyword {@code block} (or its short 583 * form {@code blk}) as in {@code block=Mongolian} or {@code blk=Mongolian}. 584 * <p> 585 * The block names supported by <code>Pattern</code> are the valid block names 586 * accepted and defined by 587 * {@link java.lang.Character.UnicodeBlock#forName(String) UnicodeBlock.forName}. 588 * <p> 589 * 590 * <b><a name="ucc">Categories</a></b> may be specified with the optional prefix {@code Is}: 591 * Both {@code \p{L}} and {@code \p{IsL}} denote the category of Unicode 592 * letters. Same as scripts and blocks, categories can also be specified 593 * by using the keyword {@code general_category} (or its short form 594 * {@code gc}) as in {@code general_category=Lu} or {@code gc=Lu}. 595 * <p> 596 * The supported categories are those of 597 * <a href="http://www.unicode.org/unicode/standard/standard.html"> 598 * <i>The Unicode Standard</i></a> in the version specified by the 599 * {@link java.lang.Character Character} class. The category names are those 600 * defined in the Standard, both normative and informative. 601 * <p> 602 * 603 * <b><a name="ubpc">Binary properties</a></b> are specified with the prefix {@code Is}, as in 604 * {@code IsAlphabetic}. The supported binary properties by <code>Pattern</code> 605 * are 606 * <ul> 607 * <li> Alphabetic 608 * <li> Ideographic 609 * <li> Letter 610 * <li> Lowercase 611 * <li> Uppercase 612 * <li> Titlecase 613 * <li> Punctuation 614 * <Li> Control 615 * <li> White_Space 616 * <li> Digit 617 * <li> Hex_Digit 618 * <li> Join_Control 619 * <li> Noncharacter_Code_Point 620 * <li> Assigned 621 * </ul> 622 * <p> 623 * The following <b>Predefined Character classes</b> and <b>POSIX character classes</b> 624 * are in conformance with the recommendation of <i>Annex C: Compatibility Properties</i> 625 * of <a href="http://www.unicode.org/reports/tr18/"><i>Unicode Regular Expression 626 * </i></a>. 627 * 628 * <table border="0" cellpadding="1" cellspacing="0" 629 * summary="predefined and posix character classes in Unicode mode"> 630 * <tr align="left"> 631 * <th align="left" id="predef_classes">Classes</th> 632 * <th align="left" id="predef_matches">Matches</th> 633 *</tr> 634 * <tr><td><tt>\p{Lower}</tt></td> 635 * <td>A lowercase character:<tt>\p{IsLowercase}</tt></td></tr> 636 * <tr><td><tt>\p{Upper}</tt></td> 637 * <td>An uppercase character:<tt>\p{IsUppercase}</tt></td></tr> 638 * <tr><td><tt>\p{ASCII}</tt></td> 639 * <td>All ASCII:<tt>[\x00-\x7F]</tt></td></tr> 640 * <tr><td><tt>\p{Alpha}</tt></td> 641 * <td>An alphabetic character:<tt>\p{IsAlphabetic}</tt></td></tr> 642 * <tr><td><tt>\p{Digit}</tt></td> 643 * <td>A decimal digit character:<tt>p{IsDigit}</tt></td></tr> 644 * <tr><td><tt>\p{Alnum}</tt></td> 645 * <td>An alphanumeric character:<tt>[\p{IsAlphabetic}\p{IsDigit}]</tt></td></tr> 646 * <tr><td><tt>\p{Punct}</tt></td> 647 * <td>A punctuation character:<tt>p{IsPunctuation}</tt></td></tr> 648 * <tr><td><tt>\p{Graph}</tt></td> 649 * <td>A visible character: <tt>[^\p{IsWhite_Space}\p{gc=Cc}\p{gc=Cs}\p{gc=Cn}]</tt></td></tr> 650 * <tr><td><tt>\p{Print}</tt></td> 651 * <td>A printable character: {@code [\p{Graph}\p{Blank}&&[^\p{Cntrl}]]}</td></tr> 652 * <tr><td><tt>\p{Blank}</tt></td> 653 * <td>A space or a tab: {@code [\p{IsWhite_Space}&&[^\p{gc=Zl}\p{gc=Zp}\x0a\x0b\x0c\x0d\x85]]}</td></tr> 654 * <tr><td><tt>\p{Cntrl}</tt></td> 655 * <td>A control character: <tt>\p{gc=Cc}</tt></td></tr> 656 * <tr><td><tt>\p{XDigit}</tt></td> 657 * <td>A hexadecimal digit: <tt>[\p{gc=Nd}\p{IsHex_Digit}]</tt></td></tr> 658 * <tr><td><tt>\p{Space}</tt></td> 659 * <td>A whitespace character:<tt>\p{IsWhite_Space}</tt></td></tr> 660 * <tr><td><tt>\d</tt></td> 661 * <td>A digit: <tt>\p{IsDigit}</tt></td></tr> 662 * <tr><td><tt>\D</tt></td> 663 * <td>A non-digit: <tt>[^\d]</tt></td></tr> 664 * <tr><td><tt>\s</tt></td> 665 * <td>A whitespace character: <tt>\p{IsWhite_Space}</tt></td></tr> 666 * <tr><td><tt>\S</tt></td> 667 * <td>A non-whitespace character: <tt>[^\s]</tt></td></tr> 668 * <tr><td><tt>\w</tt></td> 669 * <td>A word character: <tt>[\p{Alpha}\p{gc=Mn}\p{gc=Me}\p{gc=Mc}\p{Digit}\p{gc=Pc}\p{IsJoin_Control}]</tt></td></tr> 670 * <tr><td><tt>\W</tt></td> 671 * <td>A non-word character: <tt>[^\w]</tt></td></tr> 672 * </table> 673 * <p> 674 * <a name="jcc"> 675 * Categories that behave like the java.lang.Character 676 * boolean is<i>methodname</i> methods (except for the deprecated ones) are 677 * available through the same <tt>\p{</tt><i>prop</i><tt>}</tt> syntax where 678 * the specified property has the name <tt>java<i>methodname</i></tt></a>. 679 * 680 * <h3> Comparison to Perl 5 </h3> 681 * 682 * <p>The <code>Pattern</code> engine performs traditional NFA-based matching 683 * with ordered alternation as occurs in Perl 5. 684 * 685 * <p> Perl constructs not supported by this class: </p> 686 * 687 * <ul> 688 * <li><p> Predefined character classes (Unicode character) 689 * <p><tt>\X </tt>Match Unicode 690 * <a href="http://www.unicode.org/reports/tr18/#Default_Grapheme_Clusters"> 691 * <i>extended grapheme cluster</i></a> 692 * </p></li> 693 * 694 * <li><p> The backreference constructs, <tt>\g{</tt><i>n</i><tt>}</tt> for 695 * the <i>n</i><sup>th</sup><a href="#cg">capturing group</a> and 696 * <tt>\g{</tt><i>name</i><tt>}</tt> for 697 * <a href="#groupname">named-capturing group</a>. 698 * </p></li> 699 * 700 * <li><p> The named character construct, <tt>\N{</tt><i>name</i><tt>}</tt> 701 * for a Unicode character by its name. 702 * </p></li> 703 * 704 * <li><p> The conditional constructs 705 * <tt>(?(</tt><i>condition</i><tt>)</tt><i>X</i><tt>)</tt> and 706 * <tt>(?(</tt><i>condition</i><tt>)</tt><i>X</i><tt>|</tt><i>Y</i><tt>)</tt>, 707 * </p></li> 708 * 709 * <li><p> The embedded code constructs <tt>(?{</tt><i>code</i><tt>})</tt> 710 * and <tt>(??{</tt><i>code</i><tt>})</tt>,</p></li> 711 * 712 * <li><p> The embedded comment syntax <tt>(?#comment)</tt>, and </p></li> 713 * 714 * <li><p> The preprocessing operations <tt>\l</tt> <tt>\u</tt>, 715 * <tt>\L</tt>, and <tt>\U</tt>. </p></li> 716 * 717 * </ul> 718 * 719 * <p> Constructs supported by this class but not by Perl: </p> 720 * 721 * <ul> 722 * 723 * <li><p> Character-class union and intersection as described 724 * <a href="#cc">above</a>.</p></li> 725 * 726 * </ul> 727 * 728 * <p> Notable differences from Perl: </p> 729 * 730 * <ul> 731 * 732 * <li><p> In Perl, <tt>\1</tt> through <tt>\9</tt> are always interpreted 733 * as back references; a backslash-escaped number greater than <tt>9</tt> is 734 * treated as a back reference if at least that many subexpressions exist, 735 * otherwise it is interpreted, if possible, as an octal escape. In this 736 * class octal escapes must always begin with a zero. In this class, 737 * <tt>\1</tt> through <tt>\9</tt> are always interpreted as back 738 * references, and a larger number is accepted as a back reference if at 739 * least that many subexpressions exist at that point in the regular 740 * expression, otherwise the parser will drop digits until the number is 741 * smaller or equal to the existing number of groups or it is one digit. 742 * </p></li> 743 * 744 * <li><p> Perl uses the <tt>g</tt> flag to request a match that resumes 745 * where the last match left off. This functionality is provided implicitly 746 * by the {@link Matcher} class: Repeated invocations of the {@link 747 * Matcher#find find} method will resume where the last match left off, 748 * unless the matcher is reset. </p></li> 749 * 750 * <li><p> In Perl, embedded flags at the top level of an expression affect 751 * the whole expression. In this class, embedded flags always take effect 752 * at the point at which they appear, whether they are at the top level or 753 * within a group; in the latter case, flags are restored at the end of the 754 * group just as in Perl. </p></li> 755 * 756 * </ul> 757 * 758 * 759 * <p> For a more precise description of the behavior of regular expression 760 * constructs, please see <a href="http://www.oreilly.com/catalog/regex3/"> 761 * <i>Mastering Regular Expressions, 3nd Edition</i>, Jeffrey E. F. Friedl, 762 * O'Reilly and Associates, 2006.</a> 763 * </p> 764 * 765 * @see java.lang.String#split(String, int) 766 * @see java.lang.String#split(String) 767 * 768 * @author Mike McCloskey 769 * @author Mark Reinhold 770 * @author JSR-51 Expert Group 771 * @since 1.4 772 * @spec JSR-51 773 */ 774 775 public final class Pattern 776 implements java.io.Serializable 777 { 778 779 /** 780 * Regular expression modifier values. Instead of being passed as 781 * arguments, they can also be passed as inline modifiers. 782 * For example, the following statements have the same effect. 783 * <pre> 784 * RegExp r1 = RegExp.compile("abc", Pattern.I|Pattern.M); 785 * RegExp r2 = RegExp.compile("(?im)abc", 0); 786 * </pre> 787 * 788 * The flags are duplicated so that the familiar Perl match flag 789 * names are available. 790 */ 791 792 /** 793 * Enables Unix lines mode. 794 * 795 * <p> In this mode, only the <tt>'\n'</tt> line terminator is recognized 796 * in the behavior of <tt>.</tt>, <tt>^</tt>, and <tt>$</tt>. 797 * 798 * <p> Unix lines mode can also be enabled via the embedded flag 799 * expression <tt>(?d)</tt>. 800 */ 801 public static final int UNIX_LINES = 0x01; 802 803 /** 804 * Enables case-insensitive matching. 805 * 806 * <p> By default, case-insensitive matching assumes that only characters 807 * in the US-ASCII charset are being matched. Unicode-aware 808 * case-insensitive matching can be enabled by specifying the {@link 809 * #UNICODE_CASE} flag in conjunction with this flag. 810 * 811 * <p> Case-insensitive matching can also be enabled via the embedded flag 812 * expression <tt>(?i)</tt>. 813 * 814 * <p> Specifying this flag may impose a slight performance penalty. </p> 815 */ 816 public static final int CASE_INSENSITIVE = 0x02; 817 818 /** 819 * Permits whitespace and comments in pattern. 820 * 821 * <p> In this mode, whitespace is ignored, and embedded comments starting 822 * with <tt>#</tt> are ignored until the end of a line. 823 * 824 * <p> Comments mode can also be enabled via the embedded flag 825 * expression <tt>(?x)</tt>. 826 */ 827 public static final int COMMENTS = 0x04; 828 829 /** 830 * Enables multiline mode. 831 * 832 * <p> In multiline mode the expressions <tt>^</tt> and <tt>$</tt> match 833 * just after or just before, respectively, a line terminator or the end of 834 * the input sequence. By default these expressions only match at the 835 * beginning and the end of the entire input sequence. 836 * 837 * <p> Multiline mode can also be enabled via the embedded flag 838 * expression <tt>(?m)</tt>. </p> 839 */ 840 public static final int MULTILINE = 0x08; 841 842 /** 843 * Enables literal parsing of the pattern. 844 * 845 * <p> When this flag is specified then the input string that specifies 846 * the pattern is treated as a sequence of literal characters. 847 * Metacharacters or escape sequences in the input sequence will be 848 * given no special meaning. 849 * 850 * <p>The flags CASE_INSENSITIVE and UNICODE_CASE retain their impact on 851 * matching when used in conjunction with this flag. The other flags 852 * become superfluous. 853 * 854 * <p> There is no embedded flag character for enabling literal parsing. 855 * @since 1.5 856 */ 857 public static final int LITERAL = 0x10; 858 859 /** 860 * Enables dotall mode. 861 * 862 * <p> In dotall mode, the expression <tt>.</tt> matches any character, 863 * including a line terminator. By default this expression does not match 864 * line terminators. 865 * 866 * <p> Dotall mode can also be enabled via the embedded flag 867 * expression <tt>(?s)</tt>. (The <tt>s</tt> is a mnemonic for 868 * "single-line" mode, which is what this is called in Perl.) </p> 869 */ 870 public static final int DOTALL = 0x20; 871 872 /** 873 * Enables Unicode-aware case folding. 874 * 875 * <p> When this flag is specified then case-insensitive matching, when 876 * enabled by the {@link #CASE_INSENSITIVE} flag, is done in a manner 877 * consistent with the Unicode Standard. By default, case-insensitive 878 * matching assumes that only characters in the US-ASCII charset are being 879 * matched. 880 * 881 * <p> Unicode-aware case folding can also be enabled via the embedded flag 882 * expression <tt>(?u)</tt>. 883 * 884 * <p> Specifying this flag may impose a performance penalty. </p> 885 */ 886 public static final int UNICODE_CASE = 0x40; 887 888 /** 889 * Enables canonical equivalence. 890 * 891 * <p> When this flag is specified then two characters will be considered 892 * to match if, and only if, their full canonical decompositions match. 893 * The expression <tt>"a\u030A"</tt>, for example, will match the 894 * string <tt>"\u00E5"</tt> when this flag is specified. By default, 895 * matching does not take canonical equivalence into account. 896 * 897 * <p> There is no embedded flag character for enabling canonical 898 * equivalence. 899 * 900 * <p> Specifying this flag may impose a performance penalty. </p> 901 */ 902 public static final int CANON_EQ = 0x80; 903 904 // Android-changed: Android always uses unicode character classes. 905 /** 906 * Enables the Unicode version of <i>Predefined character classes</i> and 907 * <i>POSIX character classes</i> as defined by <a href="http://www.unicode.org/reports/tr18/"><i>Unicode Technical 908 * Standard #18: Unicode Regular Expression</i></a> 909 * <i>Annex C: Compatibility Properties</i>. 910 * <p> 911 * 912 * This flag has no effect on Android, unicode character classes are always 913 * used. 914 * 915 * @since 1.7 916 */ 917 public static final int UNICODE_CHARACTER_CLASS = 0x100; 918 919 /* Pattern has only two serialized components: The pattern string 920 * and the flags, which are all that is needed to recompile the pattern 921 * when it is deserialized. 922 */ 923 924 /** use serialVersionUID from Merlin b59 for interoperability */ 925 private static final long serialVersionUID = 5073258162644648461L; 926 927 /** 928 * The original regular-expression pattern string. 929 * 930 * @serial 931 */ 932 // Android-changed: reimplement matching logic natively via ICU. 933 // private String pattern; 934 private final String pattern; 935 936 /** 937 * The original pattern flags. 938 * 939 * @serial 940 */ 941 // Android-changed: reimplement matching logic natively via ICU. 942 // private int flags; 943 private final int flags; 944 945 // BEGIN Android-changed: reimplement matching logic natively via ICU. 946 // We only need some tie-ins to native memory, instead of a large number 947 // of fields on the .java side. 948 @ReachabilitySensitive 949 transient long address; 950 951 private static final NativeAllocationRegistry registry = 952 NativeAllocationRegistry.createMalloced(Pattern.class.getClassLoader(), 953 getNativeFinalizer()); 954 // END Android-changed: reimplement matching logic natively via ICU. 955 956 /** 957 * Compiles the given regular expression into a pattern. 958 * 959 * @param regex 960 * The expression to be compiled 961 * @return the given regular expression compiled into a pattern 962 * @throws PatternSyntaxException 963 * If the expression's syntax is invalid 964 */ compile(String regex)965 public static Pattern compile(String regex) { 966 return new Pattern(regex, 0); 967 } 968 969 /** 970 * Compiles the given regular expression into a pattern with the given 971 * flags. 972 * 973 * @param regex 974 * The expression to be compiled 975 * 976 * @param flags 977 * Match flags, a bit mask that may include 978 * {@link #CASE_INSENSITIVE}, {@link #MULTILINE}, {@link #DOTALL}, 979 * {@link #UNICODE_CASE}, {@link #CANON_EQ}, {@link #UNIX_LINES}, 980 * {@link #LITERAL}, {@link #UNICODE_CHARACTER_CLASS} 981 * and {@link #COMMENTS} 982 * 983 * @return the given regular expression compiled into a pattern with the given flags 984 * @throws IllegalArgumentException 985 * If bit values other than those corresponding to the defined 986 * match flags are set in <tt>flags</tt> 987 * 988 * @throws PatternSyntaxException 989 * If the expression's syntax is invalid 990 */ compile(String regex, int flags)991 public static Pattern compile(String regex, int flags) { 992 return new Pattern(regex, flags); 993 } 994 995 /** 996 * Returns the regular expression from which this pattern was compiled. 997 * 998 * @return The source of this pattern 999 */ pattern()1000 public String pattern() { 1001 return pattern; 1002 } 1003 1004 /** 1005 * <p>Returns the string representation of this pattern. This 1006 * is the regular expression from which this pattern was 1007 * compiled.</p> 1008 * 1009 * @return The string representation of this pattern 1010 * @since 1.5 1011 */ toString()1012 public String toString() { 1013 return pattern; 1014 } 1015 1016 /** 1017 * Creates a matcher that will match the given input against this pattern. 1018 * 1019 * @param input 1020 * The character sequence to be matched 1021 * 1022 * @return A new matcher for this pattern 1023 */ matcher(CharSequence input)1024 public Matcher matcher(CharSequence input) { 1025 // Android-removed: Pattern is eagerly compiled() upon construction. 1026 /* 1027 if (!compiled) { 1028 synchronized(this) { 1029 if (!compiled) 1030 compile(); 1031 } 1032 } 1033 */ 1034 Matcher m = new Matcher(this, input); 1035 return m; 1036 } 1037 1038 /** 1039 * Returns this pattern's match flags. 1040 * 1041 * @return The match flags specified when this pattern was compiled 1042 */ flags()1043 public int flags() { 1044 return flags; 1045 } 1046 1047 /** 1048 * Compiles the given regular expression and attempts to match the given 1049 * input against it. 1050 * 1051 * <p> An invocation of this convenience method of the form 1052 * 1053 * <blockquote><pre> 1054 * Pattern.matches(regex, input);</pre></blockquote> 1055 * 1056 * behaves in exactly the same way as the expression 1057 * 1058 * <blockquote><pre> 1059 * Pattern.compile(regex).matcher(input).matches()</pre></blockquote> 1060 * 1061 * <p> If a pattern is to be used multiple times, compiling it once and reusing 1062 * it will be more efficient than invoking this method each time. </p> 1063 * 1064 * @param regex 1065 * The expression to be compiled 1066 * 1067 * @param input 1068 * The character sequence to be matched 1069 * @return whether or not the regular expression matches on the input 1070 * @throws PatternSyntaxException 1071 * If the expression's syntax is invalid 1072 */ matches(String regex, CharSequence input)1073 public static boolean matches(String regex, CharSequence input) { 1074 Pattern p = Pattern.compile(regex); 1075 Matcher m = p.matcher(input); 1076 return m.matches(); 1077 } 1078 1079 // Android-changed: Adopt split() behavior change only for apps targeting API > 28. 1080 // http://b/109659282#comment7 1081 /** 1082 * Splits the given input sequence around matches of this pattern. 1083 * 1084 * <p> The array returned by this method contains each substring of the 1085 * input sequence that is terminated by another subsequence that matches 1086 * this pattern or is terminated by the end of the input sequence. The 1087 * substrings in the array are in the order in which they occur in the 1088 * input. If this pattern does not match any subsequence of the input then 1089 * the resulting array has just one element, namely the input sequence in 1090 * string form. 1091 * 1092 * <p> When there is a positive-width match at the beginning of the input 1093 * sequence then an empty leading substring is included at the beginning 1094 * of the resulting array. A zero-width match at the beginning however 1095 * can only produce such an empty leading substring for apps running on or 1096 * targeting API versions <= 28. 1097 * 1098 * <p> The <tt>limit</tt> parameter controls the number of times the 1099 * pattern is applied and therefore affects the length of the resulting 1100 * array. If the limit <i>n</i> is greater than zero then the pattern 1101 * will be applied at most <i>n</i> - 1 times, the array's 1102 * length will be no greater than <i>n</i>, and the array's last entry 1103 * will contain all input beyond the last matched delimiter. If <i>n</i> 1104 * is non-positive then the pattern will be applied as many times as 1105 * possible and the array can have any length. If <i>n</i> is zero then 1106 * the pattern will be applied as many times as possible, the array can 1107 * have any length, and trailing empty strings will be discarded. 1108 * 1109 * <p> The input <tt>"boo:and:foo"</tt>, for example, yields the following 1110 * results with these parameters: 1111 * 1112 * <blockquote><table cellpadding=1 cellspacing=0 1113 * summary="Split examples showing regex, limit, and result"> 1114 * <tr><th align="left"><i>Regex </i></th> 1115 * <th align="left"><i>Limit </i></th> 1116 * <th align="left"><i>Result </i></th></tr> 1117 * <tr><td align=center>:</td> 1118 * <td align=center>2</td> 1119 * <td><tt>{ "boo", "and:foo" }</tt></td></tr> 1120 * <tr><td align=center>:</td> 1121 * <td align=center>5</td> 1122 * <td><tt>{ "boo", "and", "foo" }</tt></td></tr> 1123 * <tr><td align=center>:</td> 1124 * <td align=center>-2</td> 1125 * <td><tt>{ "boo", "and", "foo" }</tt></td></tr> 1126 * <tr><td align=center>o</td> 1127 * <td align=center>5</td> 1128 * <td><tt>{ "b", "", ":and:f", "", "" }</tt></td></tr> 1129 * <tr><td align=center>o</td> 1130 * <td align=center>-2</td> 1131 * <td><tt>{ "b", "", ":and:f", "", "" }</tt></td></tr> 1132 * <tr><td align=center>o</td> 1133 * <td align=center>0</td> 1134 * <td><tt>{ "b", "", ":and:f" }</tt></td></tr> 1135 * </table></blockquote> 1136 * 1137 * @param input 1138 * The character sequence to be split 1139 * 1140 * @param limit 1141 * The result threshold, as described above 1142 * 1143 * @return The array of strings computed by splitting the input 1144 * around matches of this pattern 1145 */ split(CharSequence input, int limit)1146 public String[] split(CharSequence input, int limit) { 1147 // BEGIN Android-added: fastSplit() to speed up simple cases. 1148 String[] fast = fastSplit(pattern, input.toString(), limit); 1149 if (fast != null) { 1150 return fast; 1151 } 1152 // END Android-added: fastSplit() to speed up simple cases. 1153 int index = 0; 1154 boolean matchLimited = limit > 0; 1155 ArrayList<String> matchList = new ArrayList<>(); 1156 Matcher m = matcher(input); 1157 1158 // Add segments before each match found 1159 while(m.find()) { 1160 if (!matchLimited || matchList.size() < limit - 1) { 1161 if (index == 0 && index == m.start() && m.start() == m.end()) { 1162 // no empty leading substring included for zero-width match 1163 // at the beginning of the input char sequence. 1164 // BEGIN Android-changed: split() compat behavior for apps targeting <= 28. 1165 // continue; 1166 int targetSdkVersion = VMRuntime.getRuntime().getTargetSdkVersion(); 1167 if (targetSdkVersion > 28) { 1168 continue; 1169 } 1170 // END Android-changed: split() compat behavior for apps targeting <= 28. 1171 } 1172 String match = input.subSequence(index, m.start()).toString(); 1173 matchList.add(match); 1174 index = m.end(); 1175 } else if (matchList.size() == limit - 1) { // last one 1176 String match = input.subSequence(index, 1177 input.length()).toString(); 1178 matchList.add(match); 1179 index = m.end(); 1180 } 1181 } 1182 1183 // If no match was found, return this 1184 if (index == 0) 1185 return new String[] {input.toString()}; 1186 1187 // Add remaining segment 1188 if (!matchLimited || matchList.size() < limit) 1189 matchList.add(input.subSequence(index, input.length()).toString()); 1190 1191 // Construct result 1192 int resultSize = matchList.size(); 1193 if (limit == 0) 1194 while (resultSize > 0 && matchList.get(resultSize-1).equals("")) 1195 resultSize--; 1196 String[] result = new String[resultSize]; 1197 return matchList.subList(0, resultSize).toArray(result); 1198 } 1199 1200 // BEGIN Android-added: fastSplit() to speed up simple cases. 1201 private static final String FASTSPLIT_METACHARACTERS = "\\?*+[](){}^$.|"; 1202 1203 /** 1204 * Returns a result equivalent to {@code s.split(separator, limit)} if it's able 1205 * to compute it more cheaply than native impl, or null if the caller should fall back to 1206 * using native impl. 1207 * 1208 * fastpath will work if the regex is a 1209 * (1)one-char String and this character is not one of the 1210 * RegEx's meta characters ".$|()[{^?*+\\", or 1211 * (2)two-char String and the first char is the backslash and 1212 * the second is one of regEx's meta characters ".$|()[{^?*+\\". 1213 * @hide 1214 */ fastSplit(String re, String input, int limit)1215 public static String[] fastSplit(String re, String input, int limit) { 1216 // Can we do it cheaply? 1217 int len = re.length(); 1218 if (len == 0) { 1219 return null; 1220 } 1221 char ch = re.charAt(0); 1222 if (len == 1 && FASTSPLIT_METACHARACTERS.indexOf(ch) == -1) { 1223 // We're looking for a single non-metacharacter. Easy. 1224 } else if (len == 2 && ch == '\\') { 1225 // We're looking for a quoted character. 1226 // Quoted metacharacters are effectively single non-metacharacters. 1227 ch = re.charAt(1); 1228 if (FASTSPLIT_METACHARACTERS.indexOf(ch) == -1) { 1229 return null; 1230 } 1231 } else { 1232 return null; 1233 } 1234 1235 // We can do this cheaply... 1236 1237 // Unlike Perl, which considers the result of splitting the empty string to be the empty 1238 // array, Java returns an array containing the empty string. 1239 if (input.isEmpty()) { 1240 return new String[] { "" }; 1241 } 1242 1243 // Count separators 1244 int separatorCount = 0; 1245 int begin = 0; 1246 int end; 1247 while (separatorCount + 1 != limit && (end = input.indexOf(ch, begin)) != -1) { 1248 ++separatorCount; 1249 begin = end + 1; 1250 } 1251 int lastPartEnd = input.length(); 1252 if (limit == 0 && begin == lastPartEnd) { 1253 // Last part is empty for limit == 0, remove all trailing empty matches. 1254 if (separatorCount == lastPartEnd) { 1255 // Input contains only separators. 1256 return EmptyArray.STRING; 1257 } 1258 // Find the beginning of trailing separators. 1259 do { 1260 --begin; 1261 } while (input.charAt(begin - 1) == ch); 1262 // Reduce separatorCount and fix lastPartEnd. 1263 separatorCount -= input.length() - begin; 1264 lastPartEnd = begin; 1265 } 1266 1267 // Collect the result parts. 1268 String[] result = new String[separatorCount + 1]; 1269 begin = 0; 1270 for (int i = 0; i != separatorCount; ++i) { 1271 end = input.indexOf(ch, begin); 1272 result[i] = input.substring(begin, end); 1273 begin = end + 1; 1274 } 1275 // Add last part. 1276 result[separatorCount] = input.substring(begin, lastPartEnd); 1277 return result; 1278 } 1279 // END Android-added: fastSplit() to speed up simple cases. 1280 1281 /** 1282 * Splits the given input sequence around matches of this pattern. 1283 * 1284 * <p> This method works as if by invoking the two-argument {@link 1285 * #split(java.lang.CharSequence, int) split} method with the given input 1286 * sequence and a limit argument of zero. Trailing empty strings are 1287 * therefore not included in the resulting array. </p> 1288 * 1289 * <p> The input <tt>"boo:and:foo"</tt>, for example, yields the following 1290 * results with these expressions: 1291 * 1292 * <blockquote><table cellpadding=1 cellspacing=0 1293 * summary="Split examples showing regex and result"> 1294 * <tr><th align="left"><i>Regex </i></th> 1295 * <th align="left"><i>Result</i></th></tr> 1296 * <tr><td align=center>:</td> 1297 * <td><tt>{ "boo", "and", "foo" }</tt></td></tr> 1298 * <tr><td align=center>o</td> 1299 * <td><tt>{ "b", "", ":and:f" }</tt></td></tr> 1300 * </table></blockquote> 1301 * 1302 * 1303 * @param input 1304 * The character sequence to be split 1305 * 1306 * @return The array of strings computed by splitting the input 1307 * around matches of this pattern 1308 */ split(CharSequence input)1309 public String[] split(CharSequence input) { 1310 return split(input, 0); 1311 } 1312 1313 /** 1314 * Returns a literal pattern <code>String</code> for the specified 1315 * <code>String</code>. 1316 * 1317 * <p>This method produces a <code>String</code> that can be used to 1318 * create a <code>Pattern</code> that would match the string 1319 * <code>s</code> as if it were a literal pattern.</p> Metacharacters 1320 * or escape sequences in the input sequence will be given no special 1321 * meaning. 1322 * 1323 * @param s The string to be literalized 1324 * @return A literal string replacement 1325 * @since 1.5 1326 */ quote(String s)1327 public static String quote(String s) { 1328 int slashEIndex = s.indexOf("\\E"); 1329 if (slashEIndex == -1) 1330 return "\\Q" + s + "\\E"; 1331 1332 StringBuilder sb = new StringBuilder(s.length() * 2); 1333 sb.append("\\Q"); 1334 slashEIndex = 0; 1335 int current = 0; 1336 while ((slashEIndex = s.indexOf("\\E", current)) != -1) { 1337 sb.append(s.substring(current, slashEIndex)); 1338 current = slashEIndex + 2; 1339 sb.append("\\E\\\\E\\Q"); 1340 } 1341 sb.append(s.substring(current, s.length())); 1342 sb.append("\\E"); 1343 return sb.toString(); 1344 } 1345 1346 /** 1347 * Recompile the Pattern instance from a stream. The original pattern 1348 * string is read in and the object tree is recompiled from it. 1349 */ readObject(java.io.ObjectInputStream s)1350 private void readObject(java.io.ObjectInputStream s) 1351 throws java.io.IOException, ClassNotFoundException { 1352 1353 // Read in all fields 1354 s.defaultReadObject(); 1355 1356 // Android-removed: reimplement matching logic natively via ICU. 1357 // // Initialize counts 1358 // capturingGroupCount = 1; 1359 // localCount = 0; 1360 1361 // Android-changed: Pattern is eagerly compiled() upon construction. 1362 /* 1363 // if length > 0, the Pattern is lazily compiled 1364 compiled = false; 1365 if (pattern.length() == 0) { 1366 root = new Start(lastAccept); 1367 matchRoot = lastAccept; 1368 compiled = true; 1369 } 1370 */ 1371 compile(); 1372 } 1373 1374 // Android-changed: reimplement matching logic natively via ICU. 1375 // Dropped documentation reference to Start and LastNode implementation 1376 // details which do not apply on Android. 1377 /** 1378 * This private constructor is used to create all Patterns. The pattern 1379 * string and match flags are all that is needed to completely describe 1380 * a Pattern. 1381 */ Pattern(String p, int f)1382 private Pattern(String p, int f) { 1383 pattern = p; 1384 flags = f; 1385 1386 // BEGIN Android-changed: Only specific flags are supported. 1387 /* 1388 // to use UNICODE_CASE if UNICODE_CHARACTER_CLASS present 1389 if ((flags & UNICODE_CHARACTER_CLASS) != 0) 1390 flags |= UNICODE_CASE; 1391 1392 // Reset group index count 1393 capturingGroupCount = 1; 1394 localCount = 0; 1395 */ 1396 if ((f & CANON_EQ) != 0) { 1397 throw new UnsupportedOperationException("CANON_EQ flag not supported"); 1398 } 1399 int supportedFlags = CASE_INSENSITIVE | COMMENTS | DOTALL | LITERAL | MULTILINE | UNICODE_CASE | UNIX_LINES; 1400 if ((f & ~supportedFlags) != 0) { 1401 throw new IllegalArgumentException("Unsupported flags: " + (f & ~supportedFlags)); 1402 } 1403 // END Android-changed: Only specific flags are supported. 1404 1405 // BEGIN Android-removed: Pattern is eagerly compiled() upon construction. 1406 // if (pattern.length() > 0) { 1407 // END Android-removed: Pattern is eagerly compiled() upon construction. 1408 compile(); 1409 // Android-removed: reimplement matching logic natively via ICU. 1410 /* 1411 } else { 1412 root = new Start(lastAccept); 1413 matchRoot = lastAccept; 1414 } 1415 */ 1416 } 1417 1418 // BEGIN Android-changed: reimplement matching logic natively via ICU. 1419 // Use native implementation instead of > 3000 lines of helper methods. compile()1420 private void compile() throws PatternSyntaxException { 1421 if (pattern == null) { 1422 throw new NullPointerException("pattern == null"); 1423 } 1424 1425 String icuPattern = pattern; 1426 if ((flags & LITERAL) != 0) { 1427 icuPattern = quote(pattern); 1428 } 1429 1430 // These are the flags natively supported by ICU. 1431 // They even have the same value in native code. 1432 int icuFlags = flags & (CASE_INSENSITIVE | COMMENTS | MULTILINE | DOTALL | UNIX_LINES); 1433 address = compileImpl(icuPattern, icuFlags); 1434 registry.registerNativeAllocation(this, address); 1435 } 1436 compileImpl(String regex, int flags)1437 private static native long compileImpl(String regex, int flags); getNativeFinalizer()1438 private static native long getNativeFinalizer(); 1439 // END Android-changed: reimplement matching logic natively via ICU. 1440 1441 /** 1442 * Creates a predicate which can be used to match a string. 1443 * 1444 * @return The predicate which can be used for matching on a string 1445 * @since 1.8 1446 */ asPredicate()1447 public Predicate<String> asPredicate() { 1448 return s -> matcher(s).find(); 1449 } 1450 1451 /** 1452 * Creates a stream from the given input sequence around matches of this 1453 * pattern. 1454 * 1455 * <p> The stream returned by this method contains each substring of the 1456 * input sequence that is terminated by another subsequence that matches 1457 * this pattern or is terminated by the end of the input sequence. The 1458 * substrings in the stream are in the order in which they occur in the 1459 * input. Trailing empty strings will be discarded and not encountered in 1460 * the stream. 1461 * 1462 * <p> If this pattern does not match any subsequence of the input then 1463 * the resulting stream has just one element, namely the input sequence in 1464 * string form. 1465 * 1466 * <p> When there is a positive-width match at the beginning of the input 1467 * sequence then an empty leading substring is included at the beginning 1468 * of the stream. A zero-width match at the beginning however never produces 1469 * such empty leading substring. 1470 * 1471 * <p> If the input sequence is mutable, it must remain constant during the 1472 * execution of the terminal stream operation. Otherwise, the result of the 1473 * terminal stream operation is undefined. 1474 * 1475 * @param input 1476 * The character sequence to be split 1477 * 1478 * @return The stream of strings computed by splitting the input 1479 * around matches of this pattern 1480 * @see #split(CharSequence) 1481 * @since 1.8 1482 */ splitAsStream(final CharSequence input)1483 public Stream<String> splitAsStream(final CharSequence input) { 1484 class MatcherIterator implements Iterator<String> { 1485 private final Matcher matcher; 1486 // The start position of the next sub-sequence of input 1487 // when current == input.length there are no more elements 1488 private int current; 1489 // null if the next element, if any, needs to obtained 1490 private String nextElement; 1491 // > 0 if there are N next empty elements 1492 private int emptyElementCount; 1493 1494 MatcherIterator() { 1495 this.matcher = matcher(input); 1496 } 1497 1498 public String next() { 1499 if (!hasNext()) 1500 throw new NoSuchElementException(); 1501 1502 if (emptyElementCount == 0) { 1503 String n = nextElement; 1504 nextElement = null; 1505 return n; 1506 } else { 1507 emptyElementCount--; 1508 return ""; 1509 } 1510 } 1511 1512 public boolean hasNext() { 1513 if (nextElement != null || emptyElementCount > 0) 1514 return true; 1515 1516 if (current == input.length()) 1517 return false; 1518 1519 // Consume the next matching element 1520 // Count sequence of matching empty elements 1521 while (matcher.find()) { 1522 nextElement = input.subSequence(current, matcher.start()).toString(); 1523 current = matcher.end(); 1524 if (!nextElement.isEmpty()) { 1525 return true; 1526 } else if (current > 0) { // no empty leading substring for zero-width 1527 // match at the beginning of the input 1528 emptyElementCount++; 1529 } 1530 } 1531 1532 // Consume last matching element 1533 nextElement = input.subSequence(current, input.length()).toString(); 1534 current = input.length(); 1535 if (!nextElement.isEmpty()) { 1536 return true; 1537 } else { 1538 // Ignore a terminal sequence of matching empty elements 1539 emptyElementCount = 0; 1540 nextElement = null; 1541 return false; 1542 } 1543 } 1544 } 1545 return StreamSupport.stream(Spliterators.spliteratorUnknownSize( 1546 new MatcherIterator(), Spliterator.ORDERED | Spliterator.NONNULL), false); 1547 } 1548 } 1549